Gen II Prius Individual Battery Module Replacement

Thanks Brad and all of the other Priuschat members for the advice. I only did data list once, and it errored out, so I thought I was doing something wrong. I got the live data, and it confirmed the codes. Block 9 seems fine now; perhaps it took some drive time after replacement to be OK. Block 7 definitely showed that it needed replacement. Block 13 was kind of borderline. Maybe that's why it trips a code sometimes, but not others. Anyway I've order two replacment modules that are on the way.

 

Does this mean I have a speck in my eye?

Brad

 

From Kiwi's point of view, yes.

 

OK, I lied... temp sensor wire was broken off... soldered it and the other wire broke off... barely was able to solder both next to each other... PLEASE BE CAREFUL WITH THE SUPER DELICATE TEMPERATURE SENSORS!!!!

Posted via the PriusChat mobile app.

 

I think I remember telling you that this is a learning experience.

JeffD

 

Yes you did... all too true... lol thanks : p

Posted via the PriusChat mobile app.

 

My replacement modules are in. I tried load testing the two individual modules in the bad block, and I still couldn't tell which one was bad. I even tried increasing the length of the load test. The decrease in voltage was pretty consistent. Do I need to borrow a friend's hobby charger to help me identify the bad module? When doing the forced charge and discharge recommended by others, you could definitely see the bad blocks acting very differently than the others, so I know something's wrong with them.

I'm almost thinking about playing Russian Roulette, going with my 50/50 odds, and just choosing one, and seeing if it eliminates the DTC.

 

How much of a load and for how long? You need to apply the load until the voltage decreases to 6.5V. It's possible that both could be weak so you should test a module from another block as a comparison.

How were you planning to recharge the modules after the load test? You were going to need RC balance chargers or a grid charger for this anyway.

Unless you enjoy installing and removing your HV battery more than once, don't flip a coin with the modules - it's a suckers bet.
Find the weak module(s), replace them with known good ones.

 

Just to be sure. The block count starts opposite the battery ECU.

Brad

 

I used the Harbor Freight 6/12V Battery Load Tester, which S Keith had recommended in another post. The Schumacher version of the battery tester says it uses 50 amps for a 6V battery and 100 amps for a 12V battery. I'm not sure how the tester knows whether to use 50 amps or 100 amps. Anyway, I applied load for 15 seconds, whereas last week I applied load for only 8 seconds. My resting voltages were all between 7.66 to 7.69. I load tested 14 of the modules, and they all dropped to 6.70 to 6.81. The one that dropped to 6.70 is sort of an outlier, the rest were really 6.75 to 6.81. When using a stopwatch, the voltages can move a few hundredths in a moment.

I'll redo the test so it decreases to 6.5V. What kind of differing behavior should I see at this point? Is 6.5V a good point, where you identify a module with one cell failing out of the six?

Regarding replacing the modules, I had two bad blocks identified through Techstream. I only bought 2 replacement modules instead of 4.

 

You need to use a stop-watch to record the time it takes each module to reach your selected voltage (6.5 V?). The failed modules become apparent as they are the ones that fall to the voltage the fastest. This assumes that the majority of modules are ok, and so the outliers stand out–which should be the case as you have bad block DTCs.

 

I own the grid charger.

OK, that sounds good. Slightly different test, but similar in concept. Instead of keeping a constant amount of time and measuring the drop in voltage, I'll be keeping the drop in voltage constant (assuming their starting voltage are roughly equal) and measuring the time.

 

Correct, the starting voltage would have to be the same. Your original way works too, and is probably more accurate. Note the original voltage and the ending voltage (while still under load) and record. Then calculate the difference. Just be sure to use a good amount of time so the majority of modules are nearing 6 - 6.5 V. The bad blocks have the largest difference.

 

Resting voltage is a really poor indicator of SOC. Modules can be nearly empty and still have close to the same resting voltage they did full. So if you compare the load test results of a nearly full and nearly empty module it's meaningless gibberish. Use your grid charger to get all the modules to 100% SOC. Then run the load test.

Brad

 

I read somewhere else where S Keith recommended grid charging for 24 hours then resting for 24 hours before a load test? Do you think that's necessary?

Could I force charge the pack to get part of the way to 100% then grid charge to shorten that time span?

 

I left my Prius sitting without running it for some time, at least a week and then pulled the battery and measured the voltage of each module with my multi-meter. I found that most modules measured 7.69 or 7.70 Volts. One was 6.57, obviously bad. The remainder measured 7.61 to 7.67 volts. Would I be wrong to assume that the remainder had worse internal resistance than the other good modules? I don't want to buy a grid charger and the Hitec hobby charger shows the "remainder" modules have significantly worse discharge mAh than the "good" modules.

I understand why using the grid charger gives a more accurate picture of the internal resistance or load bearing capacity of the modules. Two questions:

Can I use the Prius' to charge the modules instead of the grid charger even if the red triangle shows up?

Is there any correlation between the apparent ability of a hobby charger to restore module mAh and the actual use of the modules with high charge and discharge rates in the Prius?

 

Hi everyone just a quick question for anyone using the old indi chargers (16x5 or 16x7) have a few of these solid old beasties and i want to get them in sync with one and other, the firmware on these units does not cater for user adjustable voltage calibration.

I see one internal trimpot, don't want to spin it unless I knows its the base voltage adjust, does anyone know?

Cheers .. Matt

 

Ran the load test as you said after charging the battery. As a side note my SOC never reached 100%. After testing I still could not see any discernible differences between the two modules in the blocks that had codes. Did the load test both ways, for a set time and also measuring time needed to get down to a certain voltage. I did not see any severe voltage drops, and the voltage drops in the bad blocks were almost identical to good blocks.

I had force charged the car a few nights ago and didn't see any erratic behavior in the bad blocks like before. I told myself I'm just going to put everything back and drive this car for 30 minutes and see what happens. Drove 30 minutes code-free, whereas my car would previously throw codes 10 to 15 minutes after clearing codes.

I'm partially kicking myself if this was the fix (and not considering it earlier) but partially thrilled if everything is fine now. I had a good amount of bus bar corrosion that I cleaned off. Nuts were ok and I couldn't see any leaks, just the copper was super dirty.

 

What did you do to bring all the modules to the same SOC before reassembling the battery? Running different load tests could spread the SOC apart.

Brad

 

I didn't do anything, and I could definitely see the different SOC before I set out on the road. Last night's late drive was just meant as a "What the heck, just see what happens" experiment.

I'm going to do another test drive this evening, and then start a three cycle charge-discharge with the grid charger.